Achieving Efficient Light‐Emitting Diodes by Controlling Phase Distribution of Quasi‐2D Perovskites

Quasi‐2D perovskite light‐emitting diodes (PeLEDs) are promising candidates to realize superior luminescent. However, the poorly controlled phase distribution and surface defects hinder the improvement of the device's performance. Here, by introducing rubidium bromide (RbBr) to tune the crystallization kinetics of quasi‐2D perovskites, more uniform phase distribution is achieved through the suppression of medium‐n phases, resulting in narrower emission spectrum and more efficient energy transfer. Meanwhile, the defects are effectively passivated by the addition of RbBr. As a result, the photoluminescence quantum yield (PLQY) of quasi‐2D perovskite films increases significantly from 45.6% to 81.3%, and the maximum external quantum efficiency (EQE) of PeLEDs reaches 18.92%. This finding provides a new insight into the phase distribution control of quasi‐2D perovskites and the further improvement of PeLEDs.

tures multiple phases rather than a single phase because the formation energies for different quasi-2D phases are quite similar. [21][22][23] The poorly controlled phase distribution results in spatial and energetic disorder, which leads to an inefficient energy funnel and increases energy loss during the transfer process. [9,[24][25][26][27] Meanwhile, the reduced dimension leads to abundant grain boundaries and surface defects, which may introduce more non-radiative recombination. [28][29][30] Therefore, optimizing the distribution of different n phases and passivating the defects are both meaningful to realize high-quality quasi-2D perovskite films and improve the luminous efficiency of corresponding PeLEDs.
Strategies have been adopted to solve these problems by introducing molecular additives. [31] For example, Ban et al. introduced a molecular additive named 1,4,7,10,13,16-hexaoxacyclooctadecane (crown) to suppresses crystallization of the organic phase and achieved more controlled phase distribution. [9,32] Ma et al. used a bifunctional molecular additive named Fluorinated Triphenylphosphine oxide (TFPPO), which inhibits the formation of the low-n phase while passivating the surface of perovskite quantum wells. [33] Besides molecular additives, metal salts have also been used to control the crystallization process of quasi-2D perovskite. For example, Kong et al. exploited mesylate (MeS) to interact with spacer butylammonium (BA) cations, which resulted in an increased proportion of large n phases and a concomitant decrease in defect density. [34] Guo et al. utilized alkali-metal bromides to manipulate the nucleation and growth of quasi-2D perovskites, resulting in a spatially more homogeneous distribution of different-n phases Quasi-2D perovskite light-emitting diodes (PeLEDs) are promising candidates to realize superior luminescent. However, the poorly controlled phase distribution and surface defects hinder the improvement of the device's performance. Here, by introducing rubidium bromide (RbBr) to tune the crystallization kinetics of quasi-2D perovskites, more uniform phase distribution is achieved through the suppression of medium-n phases, resulting in narrower emission spectrum and more efficient energy transfer. Meanwhile, the defects are effectively passivated by the addition of RbBr. As a result, the photoluminescence quantum yield (PLQY) of quasi-2D perovskite films increases significantly from 45.6% to 81.3%, and the maximum external quantum efficiency (EQE) of PeLEDs reaches 18.92%. This finding provides a new insight into the phase distribution control of quasi-2D perovskites and the further improvement of PeLEDs.

Introduction
Lead halogen perovskite is a promising light-emitting material owing to its low cost, solution processability, tunable bandgap, high light-emitting efficiency, and color purity. [1][2][3][4] Due to the rapid progress in aspect such as defect passivation, the external quantum efficiencies (EQE) of perovskite light-emitting diodes (PeLEDs) have exceeded 20% for near infrared and red colors and 25% for green color, which exhibit their great potential www.advelectronicmat.de and promoted energy transfer. [25] So far, the general strategy to improve energy transfer efficiency is by eliminating the low-n phases to obtain more uniform phase distribution.
In this work, we found that reasonable control of the growth of low-n phases is also conducive to improve the luminous efficiency of quasi-2D perovskites. By introducing rubidium bromide (RbBr) to tune the crystallization kinetics of quasi-2D perovskites, the phase distribution was mainly concentrated in n = 2 phase in the beginning and continued to grow from low-n to high-n phases after annealing. As a result, the final phase distribution was mainly concentrated in n = 2 and high-n phases. The significant reduction of the medium-n phases made the phase distribution more uniform, resulting in narrower emission spectrum, less nonradiative recombination, and more efficient energy transfer. Meanwhile, the addition of RbBr effectively passivated the defects of quasi-2D perovskites. As a result, the photoluminescence quantum yield (PLQY) of quasi-2D perovskite films increased significantly from 45.6% to 81.3%, which is helpful to improve the performance of the device. Finally, we prepared a quasi-2D PeLEDs with a maximum EQE of 18.92%. Different from previous work, we have successfully demonstrated that promoting the growth of low-n phase (mainly n = 2) and thus suppressing the medium-n phases with Rb + addition is also an alternative way to obtain more homogeneous phase distribution and highly efficient quasi-2D PeLEDs. Our work provides a new insight into the crystallization control of quasi-2D perovskite and the further improvement of device performance.

Effect of Rb + to the Nucleation and Crystallization of Quasi-2D Perovskites
In order to understand the effect of additional RbBr, the nucleation and crystallization process of quasi-2D perovskite were analyzed by a series of characterizations. To prepare the perovskite films, alkali metal ions and other perovskite precursors were first mixed and dissolved together in dimethyl sulfoxide (DMSO) to obtain the perovskite precursor solution. [25,35] Chlorobenzene (CB) was used as the anti-solvent during the spin coating process, so that the perovskite precursors were rapidly supersaturated and crystallized to form high-quality perovskite films. [31] In order to conveniently express the amount of additional alkali metal ions, it is expressed by "wxABr," where A represents different alkali metals (Na, K, Rb, Cs), and x represents the molar ratio between ABr and FABr. [25,36] To understand the influence of RbBr additive to the phase distribution of quasi-2D perovskite films, we first investigate the absorption spectrum of the perovskite films. As shown in Figure 1a, different absorption peaks appear in the absorption spectrum, which corresponds to different n-phases. For the pristine quasi-2D perovskite film, absorption peaks at 433, 464, 493. and 532 nm were observed before annealing, which corresponds to n = 2, 3, 4, and high-n phase respectively. [25,31] Due to the well-known energy funnel effect of quasi-2D perovskites, in which the energy of the low-n phases will flow to the Figure 1. Effect of Rb + on the phase distribution of quasi-2D perovskites. a) Absorption spectra of quasi-2D perovskite films with and without addition of RbBr (0.5 molar ratio between Rb + and FA + ) in annealed and unannealed conditions. b, c) Absorption spectra (b) and XRD patterns (c) of quasi-2D perovskite films with different concentrations of RbBr doping after annealing. d) Schematic illustration of the crystallization process of quasi-2D perovskites with Rb + doping.

www.advelectronicmat.de
high-n phases autonomously and finally concentrated in the narrowest bandgap phase, resulting in the high-n phase luminescence. Films with different n-phase distributions may emit light of different colors under UV light. [24] As a result, the pristine film emitted green light as reported elsewhere. [31] While for Rb-added film, the emitted color is blue before annealing, as shown in Figure S1, Supporting Information. This can be explained by the absorption spectrum that both the medium-n phases (n = 3, 4) and high-n phases (n > = 5) were significantly suppressed for Rb-added film, resulting in the phase distribution concentrated in low-n phase (n = 2). After the subsequent annealing process, the pristine film still emitted green light while the color of Rb-added film changed from blue to green. Accordingly, the absorption spectrums show that the phase distribution kept almost unchanged for the pristine film, while the high-n phases started to appear in the Rb-added film after annealing. This phenomenon can be explained by the growth of low-n phases which transformed into high-n phases in the annealing process for Rb-added film. [37] We further studied the effect of different doping ratios of RbBr to the phase distribution. As shown in Figure 1b and Table S1, Supporting Information, as the doping ratio of RbBr increases, the phase distribution tends to more concentrate to low-n phases (mainly n = 2) in the perovskite films. This is further confirmed by the X-ray diffraction (XRD) measurement that the (060) peak which corresponds to n = 2 phase is significantly enhanced with the doping ratio of RbBr, as shown in Figure 1c. [31,38] These findings indicate that the growth of n = 2 phase can be significantly enhanced by the addition of RbBr. With a proper doping ratio of RbBr, the phase distribution of perovskite film can be mainly concentrated in n = 2 and high-n phases after annealing, with significantly reduction of mediumn phases. According to previous reports, the multi-phase distribution of quasi-2D perovskite films will lead to serious energy disorder, resulting in non-radiative recombination and degradation of the device performance. [9] Therefore we can conclude that, with the help of RbBr which can promote the growth of n = 2 phase and on the other hand inhibit the growth of medium-n phases, more homogeneous phase distribution could be obtained, which is conducive to the energy transfer and will be discussed in the following sections.
To better understand the role of Rb + in the crystallization process, we systematically investigated the influence of different alkali metals additives to the crystallization kinetics of quasi-2D perovskite films. According to the previous report, the enhanced Coulomb interaction between alkali-metal cations and negatively charged PbBr 6 4obey the trend of K + < Na + < Li + as the ionic radius decrease, which will hinder the crystal growth and lead to more low-n phases. [25] In our experiment, the Na-added and the K-added, pristine films immediately emit green light after spin coating. The absorption spectrum shows that more low-n phases appear in Na-added film including n = 1 phase, which is consistent with the previous report. [25] According to this trend, the low-n phases will be further reduced in Rb + and Cs + cases. However, our experiments showed different results with more low-n phases than expected in Rb-added films as described above. For Cs-added film, more low-n phases appeared including n = 1 phase even compared to Na-added film, as shown in Figure S1, Supporting Information. Furthermore, the Cs-added film still emit blue light after annealing, which exhibits stronger confinement effect to low-n phases than films with other alkali metals additives. The abnormal phenomenon shows that there may be some other different mechanism works in Rb + and Cs + cases.
The n-phase distribution and emission color of the PEABradded film was similar to those with CsBr addition. According to the previous report, alkali metals can also act as spacer and participate in the formation of quasi-2D phases. [35,39,40] These results indicate that the role of additional Cs + and Rb + may be similar as PEA + , which acts as spacer in forming low-n phases. In addition, we investigated the addition of excess RbBr and CsBr in 3D perovskite solutions as shown in Figure S3, Supporting Information. We found that a low-n phase was generated after the addition of RbBr and CsBr, indicating that Rb+ and Cs+ can indeed act as spacers and participate in the formation of the quasi-2D phase. X-ray photoelectron spectroscopy (XPS) measurements were conducted to understand whether Rb + will enter the lattice to change the structure of the perovskite. The results are shown in Figure S4, Supporting Information, which shows the presence of Rb, Pb, Br, and N elements in the perovskite film, and the presence of Rb indicates that Rb + participates in the crystallization of perovskite. The XPS spectra of Pb 4f at the peaks of 138.6 and 143.4 eV and Rb 3d at the peaks of 110.0 eV both did not move, which indicates that Rb + did not affect the internal structure of the perovskite crystal. [35] Considering the stronger steric effect as the ionic radius of alkali metals increase, we speculate that the Cs + is too big and heavy to move once it participates in the formation of low-n phases, leading to the most low-n phases including n = 1 phase compare to other alkali metal ions. [41] While Rb + plays a similar role as Cs + , its smaller ionic radius and weaker structural stability in the perovskite crystal make it easier to move especially during annealing process. [40,42,43] It is worth noting that n = 1 phase is unfavorable to luminous performance due to the strong exciton-phonon coupling which will lead to severe nonradiative recombination. [44,45] Therefore, Rb + is a better choice than larger spacers such as Cs + and PEA + as the additive to control the phase distribution of quasi-2D perovskite film. Exciton binding energy is a parameter influencing the recombination mode (excitonic or bimolecular recombination) and is related with the recombination rate closely. We measured the exciton binding energy through temperature-dependent PL ( Figure S5, Supporting Information). Through curve fitting of the integrated PL intensity versus temperature. Exciton binding energy increases with the addition of RbBr, especially at 0.25 RbBr. This is because RbBr promotes the formation of low n phases. [25] Based on the above discussion, we try to illustrate the nucleation and crystallization process of quasi-2D perovskites with the addition of RbBr. As shown in Figure 1d, the additional Rb + , which is uniformly dispersed in the precursor, acts as spacer in the nucleation process of quasi-2D perovskites at first. Therefore, a large number of low-n phases (mainly n = 2) are initially formed due to the limiting effect of Rb + on the crystal growth. However, some of the Rb + ions may fall off during the annealing process, owing to the weak combination between Rb + and Br − as well as the instability of the corresponding perovskite crystal structure. [25,35] Meanwhile, the residual solvent www.advelectronicmat.de further volatilizes which leads to the continuous growth of the perovskite crystal. As a result, some of the low-n phases grow into high-n phases at last. Taking the advantage of Rb + , we can obtain quasi-2D perovskite films mainly concentrated in low-n phases (n = 2) and high-n phases with few medium-n phases. The more homogenous phase distribution can effectively suppress the non-radiative recombination caused by energetic disorder, which will be discussed in the following section.

Energy Transfer Kinetics of Quasi-2D Perovskites with Rb + Doping
Transient absorption (TA) measurements were performed to study the kinetics of energy transfer in quasi-2D perovskites. [13,20,46] As shown in Figure 2a,b, and Figure S6, Supporting Information, both pristine and Rb-added perovskites have three prominent photobleaching (PB) peaks, corresponding to n = 2, n = 3, and high-n phases, respectively. Compared with the absorption spectrum, the PB peak of n = 4 phase in the TA spectrum cannot be clearly distinguished, probably due to the close packing of the PB peaks and strong excited state absorption. [47] The evolution of photocarriers in quasi-2D perovskite films can be studied by the variation curve of the intensity of the PB peaks. At 0.1 ps, there is only PB peak of n = 2 phase, while the PB peak of higher-n phases starts to appear at 1 ps. That means the energy transfers from low-n phases to high-n phases within 1 ps, which indicates that the energy transfer in the quasi-2D perovskite film is quite fast. [25] To further investigate the kinetics of energy transfer, a multi-curve fit was performed on the TA signal, which were fitted at specific wavelengths corresponding to the PB peaks of n = 2, n = 3, and n = 5. The fitting results are listed in Table S2, Supporting Information. [34,48] As shown in Figure 2c,d, by comparing the pristine and Rb-added perovskite films, it is found that the time constants (τ1) of the ultrafast decay component of Rb-added film are 172 fs (n = 2) and 225 fs (n = 3), which is better matched to the formation (rise) time constant (τet) of 846 fs for n = 5 PB peak. In addition, the formation time of n = 5 PB peak in the perovskite films were further compared. The formation (rise) time constant (τet) of Rb-added film is 846 fs, which is shorter than 1136 fs for the pristine film. This result indicates that the addition of RbBr accelerates the energy transfer from low-n to high-n phases, which will lead to more efficient radiative recombination.

Characterization of Quasi-2D Perovskite films with RbBr Doping
To fully understand the effect of RbBr addition on aligning 2D perovskites, the surface topography of the films was first characterized. As the scanning electron microscopy (SEM) shown in Figure 3a and Figure S7, Supporting Information, the pristine films have a corrugated structure, which is consistent with the previously report. With the addition of RbBr, the corrugation of the film is greatly reduced which becomes more dense and smooth. [41] As the atomic force microscopy (AFM) images shown in Figure S8, Supporting Information, the root mean square of the film is reduced from 7.53 to 2.55 nm, indicating a reduction in the roughness of the perovskite film with the increasement of RbBr addition. When the doping ratio of RbBr increases to 0.75, some tiny particles were observed on the surface of the RbBr-added film, which may be detrimental to device performance. [49]

www.advelectronicmat.de
Photoluminescence (PL) measurements were also performed on quasi-2D perovskite films to study their luminescence properties. As shown in Figure 3b, the luminescence intensity of PL is enhanced with the addition of RbBr. This is consistent with the significant increasement of PLQY from 45.76% to 81.3%, as shown in Figure 3c and Figure S9, Supporting Information, indicating that the addition of RbBr can effectively passivate the defects and suppress the nonradiative recombination. Besides, for both films with and without the addition of RbBr, there is only one main luminescent peak around 530 nm, which is attributed to the energy transfer from low-n phases to high-n phases. [25] To analyze the details in a shorter wavelength range, the normalized PL spectrum of the films is drawn in logarithmic coordinate as shown in Figure 3d. It can be seen that there are also luminescent peaks at ≈430 and 460 nm, corresponding to n = 2 and n = 3 phases in the pristine film, which indicates that there is a certain degree of energy waste in the luminescence process. Interestingly, the addition of RbBr reduces the luminescence intensity of n = 3 and even n = 2 phases, which is opposite to the increase of n = 2 phase as the absorption spectrum and XRD shown in Figure 1c. This phenomenon can be attributed to the more efficient energy transfer due to the reduction of medium-n phases and thus more homogenous phase distribution with the addition of RbBr, which is consistent with the result of TA measurements. In addition, as shown in Figure 3e, the full width at half maxima (FWHM) of PL decreases from 27 to 22 nm after the addition of RbBr. The reduced FWHM of PL further indicates that the as-prepared perovskite films have a more homogeneous phase distribution. Meanwhile, the more homogeneous phase distribution also improves the uniformity of the perovskite films, which can be confirmed by the PL spectra of different selected regions as shown in Figure S10, Supporting Information. [50] To further investigate the radiative recombination properties of quasi-2D perovskite films, time-resolved photoluminescence (TRPL) measurements were performed. As shown in Figure 3f and Table S3, Supporting Information, the PL lifetime of the pristine film is 139.80 ns, which increases to 441.20 ns as the doping ratio of RbBr increases to 0.75. The prolonged PL lifetime further confirms that the addition of RbBr can effectively passivate the defects and suppress the non-radiative recombination, which is consistent with the result of PLQY. [51]

Properties of Quasi-2D Perovskite Light-Emitting Diodes
Charge balance is critical to device performance. According to previous reports, the transport capacity of electrons is stronger than that of holes. [25,31] Therefore, we prepared single holes and single electronic devices with ITO/PEDOT:PSS/Perovskite/Ag and ITO/SnO 2 /Perovskite/TPBi/LiF/Al structures for verification. As shown in Figure S11, Supporting Information, the electron transport capability of the device is indeed much stronger than the transport capacity of holes. We doped PSS-Na into PEDOT: PSS to increase the work function, which can improve hole injection. The quasi-2D PeLEDs with the device structure of ITO/m-PEDOT:PSS/Perovskite/ TPBi/LiF/ Al were fabricated, as shown in Figure 4a. Figure 4b shows the cross-sectional SEM image of the device, and the thicknesses of each functional layers are m-PEDOT:PSS, quasi-2D perovskite, TPBi, LiF/Al electrode. [31] Figure 4c is the energy

www.advelectronicmat.de
level arrangement corresponding to each functional layers of the device, in which the energy level of quasi-2D perovskite was tested by ultraviolet photoelectron spectroscopy (UPS). [25] It was found that the work function of the quasi-2D perovskite film was basically unchanged after the addition of RbBr, as shown in Figure S12, Supporting Information. The EL spectrums of PeLEDs are shown in Figure 4d, where the EL peak appears at about 530 nm and the FWHM decreases from 23 to 20 nm as the addition of RbBr increases. The narrower FWHM, which is consistent with the PL spectrums, again indicates that the RbBr-added perovskite films have more homogeneous phase distribution.
Device performances were further characterized to study the effect of RbBr addition. Figure 4e shows the relationship between current efficiency (CE) and voltage (V) of PeLEDs with different doping ratios of RbBr. It can be seen that the CE of PeLEDs can be significantly improved and almost doubled from 49.16 cdA −1 to 94.46 cdA −1 , as the doping ratio of RbBr increases from 0 to 0.5. When further increasing the doping ratio of RbBr to 0.75, the CE decreases to 65.66 cdA −1 . This is probably due to the excess RbBr and low-n phases, which may hinder the carrier injection and is detrimental to the device performance. Figure 4f shows the relationship between the current density (J), voltage (V), and luminance (L) of the pristine and 0.5RbBr-added PeLEDs. It can be seen that the addition of RbBr can significantly improve the luminance of the device from 4195 cdm -2 to 9207 cdm -2 . Accordingly, the EQE of PeLEDs increases from 12.71% to 18.92%, as shown in Figure 4g,h and Table S4, Supporting Information. Finally, the histogram of the maximum EQE measured from 50 devices with 0.5RbBr-added PeLEDs is shown in Figure 4i, indicating that our devices have good reproducibility, with an average EQE of 16.88% and a relative standard deviation of 4.45%. Figure S13, Supporting Information shows the electrofluorescence(EL) of the RbBr-added PeLEDs under different bias voltages. The operational stability of PeLEDs at a constant current density of 0.25 mA cm -2 with an initial luminance (L 0 ) of approximately 100 cd m -2 is shown in Figure S14, Supporting Information. The luminance of the Rb-added device decayed to half of L0 with a T 50 of 39 min, longer than the pristine device of 11.5 min. We attribute the improved stability mainly to the passivation effect that reduces ion-migration channels under an external electric field, as well as the more efficient energy transfer with less Joule heat.

Conclusion
In summary, we systematically studied the effect of RbBr addition on the crystallization of quasi-2D perovskite and the performance of corresponding PeLEDs. It was found that Rb + may act as spacer in the nucleation process to promote the growth of low-n phases (mainly n = 2), while the instability of the combination between Rb + and Br − as well as the corresponding perovskite crystal structure lead to the continuous growth of crystal and the forming of high-n phases after annealing. With the help of RbBr addition, medium-n phases were significantly reduced and more homogeneous phase distribution was obtained, which can suppress the non-radiative recombination caused by the energetic disorder. Besides, the addition of RbBr made the perovskite film more uniform in space, leading to faster and more efficient energy transfer, which is conducive to reduce the energy loss in the transfer process. Meanwhile, the addition of RbBr also effectively passivated the defects on the crystal surface, resulting in significantly increasement of the PLQY of quasi-2D perovskite films from 45.6% to 81.3%. Finally, with systematic optimization, we realized high-performance PeLEDs with a maximum luminance of 9207 cdm −2 and a maximum EQE of 18.92%. Our work paves a new way to the phase distribution control of quasi-2D perovskites and the realization of high-performance PeLEDs.

Experimental Section
Materials: All chemicals used in this work are commercially available and used without further purification. Lead bromide (PbBr 2 ), RbBr, cesium bromide (CsBr), PSS-Na (Sodium poly(4-styrenesulfonate) solution, average Mw ∼70000, 30 wt.% in H 2 O), dimethyl sulfoxide (DMSO) and chlorobenzene (CB) were purchased from Sigma-Aldrich. Device Fabrication: The ITO substrates were sequentially cleaned by sonication in deionized water, acetone, and isopropyl alcohol. The m-PEDOT:PSS aqueous solution was spin-coated onto the ITO substrates at 4500 rpm for 40 s and annealed at 150 °C for 15 min in ambient air. Then the perovskite films were prepared in a N 2 glovebox. The perovskite precursor solution was spin-coated onto the m-PEDOT:PSS substrate for 1 min at 4500 rpm. After spin coating for 20 s, 500 µL of CB was poured onto the film for pinning the perovskite crystallization, followed by annealing on a hot plate at 80 °C for 60 min. Finally, the TPBi(55nm), LiF(1nm), and Al(120nm) electrode was deposited sequentially through a thermal evaporator. The device area was 0.04 cm 2 defined by overlapping area of ITO and Al electrode.
Characterization: Photoluminescence (PL) was measured by Hitachi F-4600 fluorescence spectrophotometer with an excitation at 360 nm. The UV-vis absorption spectra were tested with a UV/vis/NIR spectrophotometer (Shimadzu, UV-3600). The broadband femtosecond TA spectra of the perovskite films were measured using the Ultrafast System HELIOS TA spectrometer with a 365-nm laser (0.17 mW). The PLQY of perovskite films was recorded by Enlitech's commercial PLQY measurement system and excited by a 365 nm LED. X-ray diffraction (XRD) spectra were collected using a multifunctional X-ray polydiffractometer (DY1602/Empyrean). X-ray photoelectron spectroscopy (XPS) data were obtained by Axis Ultra XPS spectrometer (Kratos, U.K.) with Kα radiation of Al, and operated at 225 W. Temperature-dependent steady-state photoluminescence (PL) spectra were obtained by FLS980 (Edinburgh Instruments Ltd.), equipped with a Xe lamp, a liquid nitrogen cryostat (Oxford Instruments, OptistatDN-V), and a photomultiplier tube 12 (PMT) detector. Timeresolved PL (TRPL) measurement was collected by using a fluorescencelifetime measurement system (HORIBA scientific). The morphology was measured using a scanning electron microscope (SEM) (Ultrahigh resolution field emission scanning electron microscope, Verios G4/Verios G4) and atomic force microscopy (AFM)(Multimode 8). Ultraviolet photoelectron spectroscopy (UPS) was measured by an X-ray Photoelectron Spectrometer (ESCALAB 250). The EQEs were calculated from the luminance, the current density, and the EL emission spectra of devices. All the device characterization tests were carried out on unencapsulated PeLEDs at room temperature in ambient air.

Supporting Information
Supporting Information is available from the Wiley Online Library or from the author.